CN112084636A - Multi-train cooperative control method and device - Google Patents

Abstract

Embodiments of the present invention provide a multi-train collaborative control method and device, the method comprising: S1, establishing a train dynamics model for urban rail transit; S2, modeling an urban rail transit train control system based on vehicle-to-vehicle communication; S3, according to the dynamic model and the model of the control system, construct an optimal control objective that comprehensively considers train formation distance convergence and speed convergence; S4, based on the artificial potential field method and Kalman filter, according to the optimal control objective, Coordinate control of multiple trains. The invention can effectively reduce the train tracking interval.

Description

A kind of multi-train cooperative control method and device

technical field

The present invention relates to the field of transportation, in particular to a method and device for cooperative control of multiple trains.

Background technique

With the rapid development of economy and urbanization, urban rail transit has become the main artery of public transportation in large and medium-sized cities in China, accounting for more than 50% of the passenger volume of public transportation in super-large cities such as Beijing and Shanghai. Therefore, the super-large cities represented by Beijing, Shanghai, Guangzhou, etc. are still facing great passenger pressure. The maximum full load rate of the passenger flow at the moment even reaches more than 120%. The passenger flow of urban rail transit has the following two characteristics: one is the tidal characteristics, that is, the passenger flow into the city is large and concentrated in the morning peak, and vice versa in the evening peak; the second is the large passenger flow at the transfer station. The main means to relieve the pressure of passenger flow are to invest more trains, shorten the departure interval of trains and shorten the stop time. Taking the tidal passenger flow as an example, too many vehicles in the direction of the peak passenger flow will cause the train to be crowded in the turn-around section, and the balanced "stop at station" transportation organization mode of urban rail transit will cause the transportation capacity in the direction of small passenger flow and in the small passenger flow section. waste. Therefore, there is a big contradiction between the non-equilibrium distribution of passenger flow and the balanced transportation mode.

The key technology of urban rail transit is communication-based train control (CBTC). The train maintains a stable safety protection interval. In the moving block mode, the train can follow the two modes of hitting the "hard wall" and hitting the "soft wall" during the running process.

In the "hard wall collision" mode, the current train thinks that the preceding train is in a fixed position, and the current train takes this fixed position as a hard wall and cannot collide. In this mode, the train needs to brake with appropriate deceleration to ensure that the train is in the Safe parking in front of the hard wall.

In the "soft wall collision" mode, not only the position of the preceding vehicle but also the speed of the preceding vehicle must be considered. When the current train is running, it will consider the dynamic operating parameters of the preceding vehicle, and then adjust the deceleration to avoid collision with the preceding vehicle. purpose of safe driving.

In most urban rail transit lines, hitting a "hard wall" is the only mode used by mobile blocking. Although mobile blocking has greatly shortened the train departure interval and improved the line capacity, the train running interval in this mode is still relatively large, especially in the face of special scenarios such as tidal passenger flow, the turnover efficiency of the train and the high passenger flow direction, area The capacity requirements of the segment cannot be matched. The deep-seated reason is that in the existing train operation control mode, even in the "hit soft wall" train tracking mode, it is not the train itself that controls the train's forward decision, but the ground zone controller (Zone Controller, ZC). According to the Movement Authority (MA) generated by the position information of the preceding vehicle, the train calculates the maximum safe speed according to the preceding vehicle information covered by the MA, and formulates its own speed control strategy under the safe speed. The train cannot directly obtain the information of the preceding vehicle to decide the control strategy, so the control mechanism of the existing train operation control system causes the train operation interval to be relatively large.

SUMMARY OF THE INVENTION

The embodiment of the present invention provides a multi-train cooperative control method, which can effectively reduce the train tracking interval.

A multi-train cooperative control method, comprising:

S1, establish an urban rail transit train dynamics model;

S2, modeling the urban rail transit train control system based on vehicle-to-vehicle communication;

S3, according to the dynamic model and the model of the control system, construct an optimal control objective that comprehensively considers train formation distance convergence and speed convergence;

S4, based on the artificial potential field method and the Kalman filter, and according to the optimized control objective, perform coordinated control on multiple trains.

A multi-train cooperative control device, comprising:

Establish units and establish a dynamic model of urban rail transit trains;

Modeling unit to model the urban rail transit train control system based on vehicle-to-vehicle communication;

A construction unit, according to the dynamic model and the model of the control system, to construct an optimal control objective that comprehensively considers the train formation distance convergence and speed convergence;

The control unit, based on the artificial potential field method and Kalman filter, performs cooperative control on multiple trains according to the optimal control objective.

The invention models the train as a discrete linear time-invariant system, takes the relative distance and relative speed between trains as the constraint conditions for controlling the formation of multiple trains, and at the same time takes into account the influence of noise in the actual formation process, and introduces a Kalman filter state observer , to ensure the convergence and robustness of the potential field algorithm. The control strategy proposed by the present invention can effectively reduce the train tracking interval, and at the same time achieve flexible configuration of train resources on the line by means of train formation, which has important practical significance.

It can be seen from the technical solutions provided by the above embodiments of the present invention that in the embodiments of the present invention,

Additional aspects and advantages of the present invention will be set forth in part in the following description, which will be apparent from the following description, or may be learned by practice of the present invention.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

1 is a schematic diagram of a multi-train cooperative control method according to the present invention;

2 is a schematic diagram of a CBTC system adding a train formation mode in an application scenario of the present invention;

FIG. 3 is a schematic diagram of the workflow of the train state observer in the application scenario of the present invention.

FIG. 4 is a schematic diagram of the vehicle speed in the formation mode in the application scenario of the present invention.

FIG. 5 is a schematic diagram of the interval between adjacent trains in the formation mode in the application scenario of the present invention.

FIG. 6 is a schematic diagram of vehicle acceleration in a formation mode in an application scenario of the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, but not to be construed as a limitation of the present invention.

In order to facilitate the understanding of the embodiments of the present invention, the following will take several specific embodiments as examples for further explanation and description in conjunction with the accompanying drawings, and each embodiment does not constitute a limitation to the embodiments of the present invention.

As shown in FIG. 1, it is a multi-train cooperative control method according to the present invention, including:

S1, establish an urban rail transit train dynamics model;

S2, modeling the urban rail transit train control system based on vehicle-to-vehicle communication;

S3, according to the dynamic model and the model of the control system, construct an optimal control objective that comprehensively considers train formation distance convergence and speed convergence;

S4, based on the artificial potential field method and the Kalman filter, and according to the optimized control objective, perform coordinated control on multiple trains.

The step 1 is specifically:

The kinematic model of the train is as follows:

x[k+1]=Ax[k]+Bu[k] (1)

x[k] is the state of the train in the kth communication cycle, u[k] is the potential field value output by the potential field function, A and B are the parameter matrices respectively;

The train state x[k] is expressed as follows:

x[k]=[s _i [k],v _i [k]] ^T (2)

where s _i [k] and v _i [k] represent the position and speed of the train, respectively.

The step 2 is specifically:

In the CBTC system, vehicle-to-vehicle communication is added to realize the coexistence of the two systems of vehicle-to-vehicle communication and vehicle-to-ground communication. Trains running in formation exchange information with the control center through vehicle-to-ground communication, and exchange information with adjacent trains through vehicle-to-vehicle communication; In the running mode, other trains except the first train are added with the cooperative control module Train Cooperative Operation to make state decisions;

In the train formation control algorithm, the formation command is issued by the ground center ATS, and the command issued includes the designation of the leader and the follower. The first train in the formation is designated as the leader, and the rest of the trains in the formation are designated as followers. The first train runs as the leader and tracks the ATO curve according to the timetable, and the rest of the trains in the formation track the position and speed of the first train as followers.

The step 4 is specifically:

S41: Collect the real-time running status of the trains in the communication topology, and obtain the position and speed information of each train;

S42: Input the position and speed information of each train into the potential field function and Kalman filter;

S43: Calculate the control force u[k] for each train according to the state potential field function and the Kalman filter;

S44: apply the control force u[k] to each train;

S45: Repeat steps S41-S44 until the train runs to the destination.

The step 43 is specifically:

Step 431, the rear vehicle establishes communication with the preceding vehicle;

Step 432, the rear vehicle receives the potential field function output u[k] of the preceding vehicle;

Step 433, the rear vehicle receives the preceding vehicle y[k]; y[k] contains speed and position information;

Step 434, the rear vehicle calculates according to the dynamic mathematical model of the front vehicle

Step 435, the rear vehicle calculates according to the mathematical model of the on-board sensor of the front vehicle

收敛到y[k]；如果判断结果为是，则表示

收敛到x[k]；如果判断结果为否，则跳到步骤433；Step 436, the following vehicle is judged

Convergence to y[k]; if the judgment result is yes, it means

Convergence to x[k]; if the judgment result is no, then jump to step 433;

Step 437 , the rear vehicle uses the converged x[k] to calculate the output u[k] of the rear vehicle potential field function.

The step 432 is specifically:

The potential field function for the control of the distance between trains is expressed as follows:

Among them, X _ij is the actual running interval of the i-vehicle and j-vehicle, d _ij is the expected minimum safety interval of the two vehicles, and k _s >0 determines the coefficient of the control input; A _ij is the adjacency matrix corresponding to the communication topology of the multi-train formation system ; The variable in A _ij is a _ij to indicate the information sharing state between trains in the formation, a _ij of 1 indicates that the information link is normal, and 0 indicates that the information link is abnormal; when X _ij =d _ij , between two adjacent trains The distance control function is 0, that is, when the two trains are at the desired distance, the absolute value of the distance control function is at the global minimum value; when X _ij >d _ij , the potential function is positive, and the "attraction" between the two trains makes The distance between the two trains becomes smaller, which has the effect of getting closer; when X _ij <d _ij , the potential function is negative, and a "repulsive force" is generated between the two trains, which has the effect of pushing farther away;

The expression of the velocity control potential function is as follows:

where k _v >0 is the gain coefficient of the potential field function, V _i is the actual speed of train i, and V _j is the speed of other trains in the communication topology.

The sum of the distance potential field and the velocity potential field is the total potential field output, and the total potential field is recorded as

The present invention also provides a multi-train cooperative control device, comprising:

Establish units and establish a dynamic model of urban rail transit trains;

Modeling unit to model the urban rail transit train control system based on vehicle-to-vehicle communication;

A construction unit, according to the dynamic model and the model of the control system, to construct an optimal control objective that comprehensively considers the train formation distance convergence and speed convergence;

The control unit, based on the artificial potential field method and Kalman filter, performs cooperative control on multiple trains according to the optimal control objective.

The application scenarios of the present invention are described below.

The invention relates to a multi-train cooperative control method considering vehicle-to-vehicle communication. In the way of train-to-train communication, the collaborative control algorithm is used to replace the mechanical coupler of the train to connect the trains virtually, so as to achieve ultra-short-distance and ultra-high-density train tracking. design issues.

FIG. 2 is a schematic diagram of a CBTC system adding a train formation mode in an application scenario of the present invention; FIG. 3 is a schematic diagram of a work flow of a train state observer in an application scenario of the present invention. FIG. 4 is a schematic diagram of the vehicle speed in the formation mode in the application scenario of the present invention. FIG. 5 is a schematic diagram of the interval between adjacent trains in the formation mode in the application scenario of the present invention. FIG. 6 is a schematic diagram of vehicle acceleration in a formation mode in an application scenario of the present invention. The following description will be given in conjunction with each figure. The present invention proposes a multi-train cooperative control method based on artificial potential field method and Kalman filter, including:

S1: Establish a dynamic model of urban rail transit trains;

S2: Model the urban rail transit train control system based on vehicle-to-vehicle communication;

S3: Construct an optimal control objective that comprehensively considers train formation distance convergence and speed convergence;

S4: Design a multi-train cooperative controller based on artificial potential field method and Kalman filter. The specific control method steps are:

S41: Collect the real-time running status of the trains in the communication topology, and obtain the position and speed information of each train;

S42: Input the position and speed information of each train into the potential field function and Kalman filter;

S43: Calculate the control force u[k] for each train according to the state potential field function and the Kalman filter;

S44: apply the control force u[k] to each train;

S45: Repeat steps S41 and S44 until the train runs to the destination.

Among them, the modeling process of controlling the multi-train formation is as follows:

1. Dynamic Model of Urban Rail Transit Train

Since the train-to-vehicle communication is periodic, the train can be modeled as a discrete linear time-invariant system. The kinematic model of the train is as follows:

x[k+1]=Ax[k]+Bu[k] (1)

In the above formula, x[k] is the state of the train in the kth communication cycle, u[k] is the potential field value output by the potential field function, and A and B are the parameter matrices respectively.

In the train kinematics model, the train state includes the position and speed information of the train. The train state x[k] is expressed as follows:

x[k]=[s _i [k],v _i [k]] ^T (2)

where s _i [k] and v _i [k] represent the position and speed of the train, respectively.

2. Modeling of urban rail transit train control model based on vehicle-to-vehicle communication

Vehicle-to-vehicle communication is added to the CBTC system to realize the coexistence of two systems of vehicle-to-vehicle communication and vehicle-to-ground communication. Trains running in formation exchange information with the control center through vehicle-to-ground communication, and exchange information with adjacent trains through vehicle-to-vehicle communication. In the formation operation mode, other trains except the first train no longer calculate the ATP curve of the train according to the MA provided by the zone controller ZC, but make state decisions by adding a cooperative control module (Train Cooperative Operation, TCO). The tracking interval of the train can be More recently, the coexistence of vehicle-to-vehicle communication and vehicle-to-ground communication at the same time makes the real-time and reliability of information exchange higher. Subsequent trains can timely understand the operation of the train ahead, so as to achieve a train tracking interval that is smaller than mobile blockade. In this paper, collaborative control is introduced, and the multiple trains in the train formation mode are regarded as a system. Under the constraints of the ATS scheduling command, the common driving goal is achieved, and the requirements of consistency and rapid convergence of the running state are met, so as to ensure the operation of the train. Safety and operational efficiency.

In the train formation control algorithm, the formation command is issued by the ground center ATS, and the command issued includes the designation of the leader and the follower. The first train in the formation is designated as the leader, and the rest of the trains in the formation are designated as Followers, do not participate in the formation without receiving the formation command. The first train runs as the leader and tracks the ATO curve according to the timetable, and the rest of the trains in the formation track the position and speed of the first train as followers.

3. The optimization objectives and constraints of the multi-train formation cooperative controller.

In the multi-train formation of urban rail transit, it is usually necessary to consider the distance and speed of trains in the formation, and control the distance and speed of trains in the formation to make multiple trains complete the formation. In the constraints, the artificial potential field method is used for the control of train spacing and train speed.

For the train spacing constraint, in the process of forming a train, when the distance between the two cars is large, the two cars will attract each other. The farther the distance is, the more obvious the gravitational force will be. And the closer the distance is, the greater the repulsion force is. At this time, the trains will be far away from each other, until the distance between the two trains stabilizes to the desired value, and the distance between the two trains reaches a stable state. The potential field function for the control of the distance between trains is expressed as follows:

Among them, X _ij is the actual running interval of vehicle i and vehicle j, _{d ij} is the expected minimum safety interval of the two vehicles, and ks >0 determines the coefficient of control input. A _ij is the adjacency matrix corresponding to the communication topology of the multi-train formation system. The variable in A _ij is a _ij to indicate the information sharing state between trains in the formation, a _ij of 1 indicates that the information link is normal, and 0 indicates that the information link is abnormal. When X _ij =d _ij , the distance control function between two adjacent trains is 0, that is, when the distance between the two trains is at the desired distance, the absolute value of the distance control function is at the global minimum value; when X _ij >d _ij , the potential If the function is positive, the "attractive force" between the two trains makes the distance between the two trains smaller, which has the effect of getting closer; when X _ij <d _ij , the potential function is negative, and there is a "repulsion" between the two trains, which plays a role in Push away effect.

The speed control potential function is introduced for the train speed constraint. The purpose of the speed control potential function is to make the speed of the trains in the formation quickly reach the consistency, assist the distance control potential function, and quickly complete the multi-vehicle formation. The expression of the velocity control potential function is as follows:

where k _v >0 is the gain coefficient of the potential field function, V _i is the actual speed of train i, and V _j is the speed of other trains in the communication topology.

The sum of the distance potential field and the velocity potential field is the total potential field output, and the total potential field is recorded as

The following describes the multi-train formation state observer.

In the actual train formation process, the influence of noise on the convergence, accuracy and robustness of the algorithm should be considered in the train formation. This paper hopes to use a filtering algorithm to filter the noise to achieve the purpose of accurately predicting the position and speed of the train. The Kalman filter is an optimization estimation algorithm and a method for designing a state observer.

The following takes the formation of two trains on the main line as an example to describe the working principle of the state observer. As shown in Figure 3, there are two trains running in front and back on the main line. The trains complete the formation and the formation state is stable. After the field function outputs u[k] and u[k] are executed by the power system of the preceding vehicle, the actual state of the preceding vehicle is x[k], the state of the preceding vehicle is sent to the rear vehicle through vehicle-to-vehicle communication, and the rear vehicle receives the The state value of the preceding vehicle is y[k], and y[k] is recorded as the observation value of the following vehicle to the preceding vehicle. It has been known from the previous analysis that due to the error of the train positioning and speed measurement sensor and the existence of communication delay, the state of the preceding vehicle obtained by the rear vehicle may not be the accurate state x[k] of the preceding vehicle, which requires the rear vehicle to observe the state of the preceding vehicle. In the on-board controller of the preceding vehicle, the train formation algorithm outputs u[k], the train power system executes u[k], and the actual state of the train is x[k].

与前车实际状态x_k是一一对应的关系，于是

能够收敛到y[k]，那么就可以保证

收敛到x[k]。The purpose of the state observer is to obtain the actual real state x[k] of the train as accurately as possible, due to the ideal measurement value of the sensor.

There is a one-to-one correspondence with the actual state x _k of the preceding vehicle, so

can converge to y[k], then it is guaranteed that

converges to x[k].

Further denoting the mechanical noise as ω[k], the noise is random, these random variables do not follow a pattern, but the average properties of the noise can be derived using probability theory. Assume that the noise ω[k] obeys a Gaussian distribution with a mean value of zero and a covariance of Q, that is, ω～N(0, Q). Since the train dynamics model has two outputs, and the dimensions of position and speed are different, Q is the covariance variance matrix. Then the kinematic equation of the train containing noise is eg.

x[k]=Ax[k-1]+Bu[k]+ω[k] (6)

In the train formation mode, the formation members make control strategies based on the position and speed of other trains, but the status information such as the position and speed of other trains received by the train is also unreliable. The reason is that the train itself is positioned. And the noise existing in the speed measurement error and the vehicle-to-vehicle communication, this type of noise is recorded as μ[k], the noise obeys a Gaussian distribution with a mean of zero and a covariance of R, μ～N(0,R).

The mathematical model of the train power unit is shown in formula (2-13):

是上一周期最优状态估计。同时理想情况下列车车载传感器得到的列车状态就是列车实际状态即：in

is the optimal state estimate for the previous cycle. At the same time, ideally, the train state obtained by the on-board sensor of the train is the actual state of the train, namely:

where C is an elementary matrix. The formula for simultaneous observation is shown in (2-15):

y[k]=Cx[k]+μ[k] (9)

称为预测部分，利用前一通信周期的估算状态

以及当前列车编队算法的输出u[k]，我们将预测部分记为

称之为列车状态在本周期的预估状态值，同时将车载传感器的测量值y[k]代入方程，用y[k]更新预估状态值，此时

部分称为后验状态估计。in the above formula

called the prediction part, which uses the estimated state of the previous communication cycle

and the output u[k] of the current train formation algorithm, we denote the prediction part as

It is called the estimated state value of the train state in this cycle. At the same time, the measured value y[k] of the on-board sensor is substituted into the equation, and the estimated state value is updated with y[k]. At this time

The part is called posterior state estimation.

以及误差协方差

由于在设计中存在机械时延，造成了列车的状态预估值的不确定性，P_k表示对列车预估状态的不确定性的度量，

和P_k-1的初始值来自于初始估计值。Two processes are required for the following car to obtain the precise state information of the preceding car. The first is the prediction process. The prediction process is used to calculate the estimated value of the train state.

and the error covariance

Due to the existence of mechanical delay in the design, the uncertainty of the state estimate value of the train is caused. P _k represents the measure of the uncertainty of the estimated state of the train,

The initial values of and P _k-1 are derived from the initial estimates.

Next is the observation process: the observation process updates the train state based on the prediction results obtained in the prediction process.

为更新后的状态值，P_k为更新后的误差协方差，K_k为卡尔曼增益，卡尔曼增益在算法中不断迭代，使更新后的状态值

的误差协方差P_k最小。

is the updated state value, P _k is the updated error covariance, K _k is the Kalman gain, and the Kalman gain is continuously iterated in the algorithm to make the updated state value

The error covariance P _k is the smallest.

The beneficial effects of the present invention are as follows:

In order to ensure the safe and efficient operation of the train formation, the present invention models the train as a discrete linear time-invariant system, takes the relative distance and relative speed between trains as constraints for controlling the formation of multiple trains, and takes into account the actual formation process. To avoid the influence of noise, a Kalman filter state observer is introduced to ensure the convergence and robustness of the potential field algorithm. The control strategy proposed by the present invention can effectively reduce the train tracking interval, and at the same time achieve flexible configuration of train resources on the line by means of train formation, which has important practical significance.

In order to verify the effectiveness of the multi-train cooperative control method based on the artificial potential field method proposed in this patent, the performance of the controller is simulated and analyzed in this section.

Assuming a scenario of two stations and one section, 4 trains are formed in formation, the first train runs according to the timetable, and the remaining 3 trains are controlled by the cooperative control algorithm. The train length, changes in train mass and the effects of noise are not considered in the simulation. Considering the coexistence of vehicle-ground communication and vehicle-vehicle communication, it is assumed that all trains in the formation can realize point-to-point communication. Therefore, the communication topology correlation matrix between all trains is:

In addition, the position and speed of the trains are marked along the track direction, the initial distance between the trains is 30m, and the initial speed is 0. Then the initial position and initial speed of the train can be expressed as a matrix

Under the constraints of the train running schedule, the working conditions of the first train throughout the entire operation are: traction, inertia, and braking. The operating conditions of other trains are under the leadership and constraints of the first train, and the other three trains gradually complete the formation under the action of the collaborative control algorithm. As shown in Figure 4, the train speed changes with time. It can be seen that the speed of the first four trains is the same in the first 30s. This is because in the initial stage, both the first car and the other cars in the formation are towed at the maximum acceleration. At 30s, the working condition of the first train changes from pulling to coasting, during which the train is only subjected to The basic resistance, the rest of the trains are affected by the first train, and the working conditions vary with the first train. It can be seen that under the leader-follower control strategy, the working condition of the first train is constrained by the timetable to ensure that the train arrives at the station on time under safety constraints, and completes the on-off service for passengers, thereby ensuring the efficiency of the train's execution of plans or tasks. At the same time, during the tracking process, the train always runs under the maximum speed limit of 22m/s, which ensures driving safety. Since the purpose of the virtual formation is to ensure that the trains in the train formation run at high speed with a very small interval to realize the rapid transfer of the trains and match the change and distribution density of the passenger flow, the relative dynamic relationship between the trains is extremely important in this process. Figure 4 is a schematic diagram of the vehicle speed in the formation mode.

In the whole operation process, the distance between trains is an important indicator to measure the quality of the algorithm. Figure 5 is a schematic diagram of the distance between adjacent trains in formation mode. In Figure 5, the interval between trains is represented. From top to bottom in the figure, the interval between 1 and 2 cars, the interval between 2 and 3 cars, and the interval between cars 3 and 4 are respectively shown. The distance between trains continued to increase, and the distance between the trains continued to decrease after the first train condition changed from traction to coasting in 30s. The distance between trains tends to be stable. After 200s, the distance between trains will reach the ideal distance range and become stable. The distance between trains is 10m.

In the process of train formation, the control decision of each train is affected by the position, speed and target speed of other trains in the formation. The characterization of the train control strategy is the acceleration of the train. Therefore, Figure 6 shows the acceleration of the train in the formation mode. Starting from the acceleration to analyze the changes in the control decision-making of trains in the process of collaborative formation, we can see that the changes in acceleration are relatively obvious, which is also in line with the characteristics of real-time dynamic control of the control algorithm. At the desired speed, the state of the train is always adjusted in dynamic equilibrium.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (7)

1. a multi-train cooperative control method, is characterized in that, comprises: S1, establish an urban rail transit train dynamics model; S2, modeling the urban rail transit train control system based on vehicle-to-vehicle communication; S3, according to the dynamic model and the model of the control system, construct an optimal control objective that comprehensively considers train formation distance convergence and speed convergence; S4, based on the artificial potential field method and the Kalman filter, and according to the optimized control objective, perform coordinated control on multiple trains. 2. The method according to claim 1, wherein the step 1 is specifically: The kinematic model of the train is as follows: x[k+1]=Ax[k]+Bu[k] (1) x[k] is the state of the train in the kth communication cycle, u[k] is the potential field value output by the potential field function, A and B are the parameter matrices respectively; The train state x[k] is expressed as follows: x[k]=[s _i [k], v _i [k]] ^T (2) where s _i [k] and v _i [k] represent the position and speed of the train, respectively. 3. method according to claim 1, is characterized in that, described step 2 is specifically: In the CBTC system, vehicle-to-vehicle communication is added to realize the coexistence of the two systems of vehicle-to-vehicle communication and vehicle-to-ground communication. Trains running in formation exchange information with the control center through vehicle-to-ground communication, and exchange information with adjacent trains through vehicle-to-vehicle communication; In the running mode, other trains except the first train are added with the cooperative control module Train Cooperative Operation to make state decisions; In the train formation control algorithm, the formation command is issued by the ground center ATS, and the command issued includes the designation of the leader and the follower. The first train in the formation is designated as the leader, and the rest of the trains in the formation are designated as followers. The first train runs as the leader and tracks the ATO curve according to the timetable, and the rest of the trains in the formation track the position and speed of the first train as followers. 4. method according to claim 1, is characterized in that, described step 4 is specifically: S41: Collect the real-time running status of the trains in the communication topology, and obtain the position and speed information of each train; S42: Input the position and speed information of each train into the potential field function and Kalman filter; S43: Calculate the control force u[k] for each train according to the state potential field function and the Kalman filter; S44: apply the control force u[k] to each train; S45: Repeat steps S41-S44 until the train runs to the destination. 5. The method according to claim 4, wherein the step 43 is specifically: Step 431, the rear vehicle establishes communication with the preceding vehicle; Step 432, the rear vehicle receives the potential field function output u[k] of the preceding vehicle; Step 433, the rear vehicle receives the preceding vehicle y[k]; y[k] contains speed and position information; Step 434, the rear vehicle calculates according to the dynamic mathematical model of the front vehicle

Step 435, the rear vehicle calculates according to the mathematical model of the on-board sensor of the front vehicle

Step 436, the following vehicle is judged

Convergence to y[k]; if the judgment result is yes, it means

Convergence to x[k]; if the judgment result is no, then jump to step 433; Step 437 , the rear vehicle uses the converged x[k] to calculate the output u[k] of the rear vehicle potential field function. 6. The method according to claim 5, wherein the step 432 is specifically: The potential field function for the control of the distance between trains is expressed as follows:

Among them, X _ij is the actual running interval of the i-vehicle and j-vehicle, d _ij is the expected minimum safety interval of the two vehicles, and k _s >0 determines the coefficient of the control input; A _ij is the adjacency matrix corresponding to the communication topology of the multi-train formation system ; The variable in A _ij is a _ij to indicate the information sharing state between trains in the formation, a _ij of 1 indicates that the information link is normal, and 0 indicates that the information link is abnormal; when X _ij =d _ij , between two adjacent trains The distance control function is 0, that is, when the two trains are at the desired distance, the absolute value of the distance control function is at the global minimum value; when X _ij >d _ij , the potential function is positive, and the "attraction" between the two trains makes The distance between the two trains becomes smaller, which has the effect of getting closer; when X _ij <d _ij , the potential function is negative, and a "repulsive force" is generated between the two trains, which has the effect of pushing farther away; The expression of the velocity control potential function is as follows:

where k _v >0 is the gain coefficient of the potential field function, V _i is the actual speed of train i, and V _j is the speed of other trains in the communication topology. The sum of the distance potential field and the velocity potential field is the total potential field output, and the total potential field is recorded as

7. A multi-train cooperative control device, characterized in that, comprising: Establish units and establish a dynamic model of urban rail transit trains; Modeling unit to model the urban rail transit train control system based on vehicle-to-vehicle communication; A construction unit, according to the dynamic model and the model of the control system, to construct an optimal control objective that comprehensively considers the train formation distance convergence and speed convergence; The control unit, based on the artificial potential field method and Kalman filter, performs cooperative control on multiple trains according to the optimal control objective.

Images